Patients with the Acquired Immunodeficiency Syndrome (AIDS), caused by Human Immunodeficiency Virus Type-1 (HIV-1) infection, develop profound immunosuppression, particularly of their innate and T-helper 1-directed cellular immunity (1). The same patients may also develop dysfunction of many organ systems, including the hypothalamic-pituitary-adrenal (HPA) axis (2). During the past 15 years, numerous reports have provided evidence for alterations in the HPA axis and its target tissues in HIV-1-infected patients. AIDS has been associated with HIV-1 or saprophytic adrenalitis, adrenal dysfunction secondary to neoplastic infiltration of the adrenal cortices, and changes related to circulating cytokines known to influence the function of the HPA axis (3). Glucocorticoids, the end-products of the HPA axis that have strong anti-inflammatory effects, are also considered for reversing HIV-1-mediated depletion of circulating CD4+ lymphocytes and slowing progression to AIDS (4).
Recent development and clinical use of three different types of anti-viral drugs, nucleoside and non-nucleoside analogues acting as reverse transcriptase inhibitors, and non-peptidic viral protease inhibitors, especially the combination therapy using any three of the above drugs (termed highly active antiretroviral therapy or HAART), have dramatically improved the clinical course of AIDS patients and prolonged their lives (5-9). However, the prolongation of the life expectancy and/or the long-term use of the above antiviral agents have generated novel morbidities and complications, which influence the patients’ quality of life and add new risk factors for premature death. Central among them is the quite common AIDS-related insulin resistance and lipodystrophy syndrome (ARIRLS), which is characterized by a striking phenotype and marked metabolic disturbances that are reminiscent of Cushing syndrome (10).
In agreement with above clinical background, acquired alterations in the sensitivity of tissues to glucocorticoids were hypothesized in AIDS patients. Depending on their direction, such alterations could be categorized as resistance or hypersensitivity, and, depending on their tissue specificity, they would be classified as generalized or tissue-specific (11, 12). Here we present our understanding of these emerging concepts and discuss their possible mechanisms and relevance to HIV-1 pathogenesis.
The HPA axis, together with the systemic sympathetic and adrenomedullary systems, are the peripheral limbs of the stress system, whose main function is to maintain basal and stress-related homeostasis (3). The HPA axis interacts with the immune system, sensing inflammatory signals, such as the cytokines tumor necrosis factor (TNFa), interleukin (IL)-1 and IL-6, and modulating the activity of this system primarily via its end product, glucocorticoids.
Glucocorticoids exert profound influences on many physiologic functions by virtue of their diverse roles in growth, development, and maintenance of cardiovascular, metabolic and immune homeostasis (13, 14). Glucocorticoids also possess potent anti-inflammatory and immunosuppressive activities, which have made them invaluable therapeutic means in inflammatory and autoimmune diseases (15).
Glucocorticoids exert their effects on their target cells through the glucocorticoid receptor (GR), a ligand-specific and -dependent transcription factor, which is ubiquitously expressed in almost all tissues (12, 16). There are two 3’ splicing variants, GRa and b, from alternative use of a different terminal exon termed 9a or 9b. GRa is the classic GR, which binds to glucocorticoids and transactivates or transrepresses glucocorticoid-responsive promoters. The GRa shuttles between the cytoplasm and the nucleus. Binding of glucocorticoids to the GR causes it to dissociate from a cytoplasmic hetero-oligomer containing heat shock proteins and to translocate into the nucleus via the nuclear pore. There, ligand-bound GRa molecules bind as dimers to specific DNA enhancer sequences, the glucocorticoid responsive elements (GREs), in the promoters of glucocorticoid-responsive genes, to modulate the transcription of these genes. On the other hand, GRb does not bind glucocorticoids and functions as a dominant negative inhibitor of GRa on GRE-containing glucocorticoid-responsive promoters (17). Its physiologic role(s) are yet to determined (18, 19).
The GRE-bound GR interacts with newly described “coactivator complexes”, which possess histone acetyltransferase (HAT) activity, as well as other chromatin modulatory protein complexes, such as SWI/SNF, SMCC and TRAP/DRIP (16, 20). One family of the coactivator molecules consisting of the homologous p300 and cAMP-responsive element binding protein (CBP) may serve as signal integrators for many transcription factors from different signal transduction cascades. Another coactivator, p/CAF, originally reported as a human homologue of yeast GCN5 that interacts with p300/CBP, is also a broad coactivator with HAT activity (21). Coactivator molecules interacting preferentially with nuclear receptors have also been described (20, 22). They include members of the p160 family of proteins: steroid receptor coactivator-1 (SRC-1); TIF-II or glucocorticoid receptor interacting polypeptide-1 (GRIP-1), also called SRC-2; the p300/CBP/co-integrator-associated protein (p/CIP), ACTR or RAC3, also called SRC-3; and the riboprotein steroid receptor coactivator (SRA) (20).
These different classes of coactivator proteins form complexes by binding to each other and with the ligand-activated nuclear receptors that interact with components of the transcription machinery on the promoter regions of responsive genes. p300/CBP and the members of the p160 family of coactivators contain one or more copies of the coactivator signature motif sequence LXXLL, which is essential for the interaction with nuclear receptors (23). The receptor-coactivator complexes not only help transduce the hormonal signal to the transcription initiation complex but also loosen chromatin structure by acetylating histones through their intrinsic histone acetyltransferase activity and facilitate the binding of transcription machinery components to DNA (16, 22).
The complex system of glucocorticoid receptor signaling suggests that the glucocorticoid activity is modulated by numerous factors at the level of the peripheral tissues (16, 24). This is referred to as “sensitivity of tissues to glucocorticoids”, which determines effectiveness of glucocorticoids in peripheral tissues. Depending on its direction, decreased or increased, it is divided into two subgroups; resistance and hypersensitivity. Both states may be generalized or tissue-specific, as well as congenital or acquired. The generalized, congenital form of glucocorticoid resistance, namely the syndrome of familial or sporadic glucocorticoid resistance, was described approximately 25 years ago (25-27). It is characterized by partial, relatively well-compensated resistance of all tissues to glucocorticoids and is mostly caused by inactivating mutations of the GR gene. On the other hand, tissue-specific, acquired forms of glucocorticoid resistance/hypersensitivity have been inferred but not fully confirmed or elucidated. Such states may be limited to certain tissues, as for instance leukocytes or adipocytes, and present with deficiency or excess glucocorticoid effect manifestations from these tissues (24). Glucocorticoid resistant asthma, rheumatoid arthritis, Crohn’s disease and systemic lupus erythematosus may be glucocorticoid resistant states of components of the immune system, while central obesity with insulin resistance, carbohydrate intolerance, dyslipidemia and hypertension, may be glucocorticoid hypersensitivity states of the adipose and/or vascular tissue (26).
The adrenal gland is one of the organs frequently found affected by HIV-1 infection at autopsy (28-30). Pathologic findings are intra-adrenal inflammatory lesions with or without necrosis, thrombosis, and/or fibrosis, as well as metastases of Kaposi sarcoma. Cytomegalovirus adrenalitis is the most common cause of the adrenal insufficiency seen in AIDS patients (29, 30), while cryptococcus, mycobacteria and pneumocystis carinii also affect the adrenal glands (28, 31). The pathologic findings vary from mild focal inflammation to extensive hemorrhagic necrosis. Although several cases with extensive adrenal necrosis and profound adrenal dysfunction have been reported (32-34), infectious adrenalitis does not usually cause clinical adrenal insufficiency in most of the AIDS patients (2). However, a recent report on 60 advanced AIDS patients with less than 50 cells/ml of peripheral CD4+ lymphocyte counts demonstrated that over 25% of these patients had abnormally low and high levels of respectively serum cortisol and plasma ACTH, reduced excretion of urinary free cortisol and/or blunted responses of serum cortisol to exsogenous ACTH (35). Thirty-eight (63.3%) patients had CMV antigenemia. Furthermore, 16 patients, among 36 patients whom the authors followed up for at least one year, developed overt adrenal insufficiency and half of them were treated with corticosteroid replacement. In conclusion, pathologic findings of the adrenal glands are frequently encountered at autopsy, yet in the majority of cases these are mild and not associated with overt primary adrenal insufficiency. Presence of adrenal insufficiency, and, hence, glucocorticoid replacement therapy, however, should be considered in end-stage AIDS patients.
Because the adrenal gland is frequently affected in AIDS patients, and because the common manifestations of these patients, such as weakness, fatigue and body weight loss, mimic those of adrenal insufficiency, many studies have been performed examining the basal and reserve activity of the HPA axis (2, 36-38). The majority of publications indicate that in AIDS patients basal levels of serum cortisol and plasma ACTH are normal or slightly elevated and their circadian rhythm is preserved (31, 39-42). Elevations of serum cortisol have been reported both in the early stages of AIDS and in severely affected, terminal patients (38, 43, 44). Twenty four-hour urinary free cortisol excretion was increased depending on the severity of AIDS (45). The adrenocortical reserve on the basis of a standard ACTH test is preserved in the majority of patients, while it is reduced in advanced cases (35). Secretion of ACTH in response to CRH is blunted, especially in the terminal stage of AIDS (37, 38, 46, 47).
Imbalance of cytokine production was suggested as a cause of low responsiveness of pituitary gland to CRH (37). Significant blunting of the ACTH response in AIDS patients was also reported in the cold immersion stress test (41). Based on the above evidence, it appears that the function of the HPA axis in AIDS patients is generally preserved and it is not likely that the hypotension, hyponatremia and hypovolemia seen in AIDS patients at the end-stage of their disease is the result of the adrenal insufficiency due to dysfunction of the pituitary gland. Rather, the adrenal insufficiency reported in AIDS patients is mainly induced by specific causes that target the adrenal glands, such as infectious adrenalitis or neoplastic infiltration and is accompanied by the typical biochemical picture of Addison’s disease.
Protease inhibitors (PIs), which inhibit the activity of viral-encoded protease and are widely used as part of HAART, act as inhibitors of one of the cytochrome P450 (CYP) enzymes, CYP3A4, which is necessary to metabolize glucocorticoids into inactive forms (48). Ritonavir is the strongest suppressor of CYP3A4-mediated 6b-hydroxylation of steroids, while indinavir and nelfinavir are moderate suppressors and saquinavir is the weakest (48). Ritonavir caused full-blown Cushing syndrome in AIDS patients treated with inhaled synthetic fluticasone by extremely reducing its metabolic clearance (49-51). Thus, glucocorticoids, even applied topically, should be used with caution in patients treated with PIs. Changing ritonavir to other PIs or use of different classes of anti-viral drugs may help reduce this characteristic side effect.
Current therapeutic regimens, including HAART, have significantly prolonged life expectancy of HIV-1-infected patients and have prevented innumerable new infections (5, 6). However, these therapy regimens are expensive and their adherence rates are sometimes low (52-54). In addition, compounds used for the treatment of AIDS often have chronic toxic side effects, such as the characteristic AIDS-related insulin resistance and lipodystrophy syndrome, which will be discussed in the following section, as well as mitochondrial toxicity, lactic acidosis, hepatotoxicity and cardiomyopathy (55-60). Thus, other antiviral agents are currently explored, including inhibitors of viral integrase, and host CXCR4 and CCR5 (61-64). In addition to these compounds, which directly interfere with viral activities, immunosuppressive agents, such as glucocorticoids and cyclosporine A, have been tested in HIV-1-infected patients, as agents that may suppress HIV-1-mediated immune activation, one of the major factors for AIDS progression and reduction of peripheral CD4+ lymphocytes (4, 65-69).
Glucocorticoids have also been used empirically in HIV-1-infected patients for their correcting their fatigue and appetite loss, as well as other complications, such as Pneumocystis carinii infection, tuberculous meningitis, autoimmune-mediated disorders and HIV-1-associated malignancies like lymphomas (70-74). The synthetic glucocorticoid prednisone at 0.3-0.5 mg/kg/day successfully increased peripheral CD4+ lymphocyte counts and prevented their reduction for up to 10 years (67). It also suppressed circulating levels of TNFa and IL-6, known indicators of HIV-1-mediated host immune activation and possible causative agents for the AIDS-associated wasting syndrome (65, 75, 76). These cytokines may also participate in the replication of the HIV-1 virus by strengthening Tat-mediated activation of the HIV-1 long terminal repeat promoter via stimulation of the nuclear factor-kB (NF-kB) (77). This beneficial effect of glucocorticoids was more obvious in patients whose immune system was less damaged (67, 69). Glucocorticoids did not alter peripheral viral load in patients who had been already treated with anti-viral drugs and thus had low viral load before the initiation of thereapy (67-69). However, one case report indicated that high doses of prednisone (100 mg for 9 consecutive days) demonstrated extremely strong suppression on the circulating virus titer of a patient infected with multi-drug-resistant HIV-1 (78). The synthetic glucocorticoid dexamethasone also inhibited elimination of CD4+ lymphocytes by macrophages isolated from HIV-1-infected patients in vitro (79). Thus, at treatment-naïve or equivalent states, glucocorticoids appear to inhibit viral replication by suppressing HIV-1-mediated inflammation and subsequent production of inflammatory cytokines. However, glucocorticoids are also risk factors for AIDS-associated metabolic complications, including sarcopenia, osteoporosis and/or osteonecrosis of the hip, and are reported to accelerate development of Kaposi’s sarcoma in the patients with pleural tuberculosis (80-84).
Thus, the therapeutic use of glucocorticoids in AIDS patients appears to be quite limited by several factors. Selective glucocorticoids or other non-steroidal compounds, with immunosuppressive actions but not metabolic side effect might be beneficial in the treatment of AIDS patients. Indeed, some of such compounds (e.g., Compound Abbott-Ligand (AL)-438, ZK216348 and the hydroxyl phenyl aziridine precursor analogue Compound A) are under investigation for their selective glucocorticoid effects (please see the Chapter “Glucocorticoid Receptor”) (85-87).
Norbiato et al. reported a distinct subgroup of AIDS patients who showed apparent adrenal insufficiency with fatigue, weakness, body weight loss, hypotension, and skin and mucosal hyperpigmentation associated with markedly elevated levels of serum cortisol and moderately increased levels of plasma ACTH (88). In these patients, the affinity of the GR to its ligand was markedly decreased in peripheral leukocytes with concurrent elevations of receptor number, suggesting that the apparent adrenal insufficiency seen in these patients might be caused by decreased sensitivity of peripheral tissues to glucocorticoids. Similar patients were subsequently reported by another group (89). Norbiato et al. estimated that up to 17 % of AIDS patients are likely to have altered glucocorticoid receptors (90), however, the accurate prevalence and mechanisms of this pathologic condition are yet to be determined.
A similar glucocorticoid resistance state was reported in glucocorticoid resistant asthma type 2 patients. In these patients, resistance to glucocorticoids was associated with reduced affinity of the GR for its ligand in limited tissues, such as peripheral leukocytes, a change that progressively reverted to normal when cells were incubated ex vivo (91). Since incubation of patients’ peripheral lymphocytes with IL-2 and IL-4 could preserve the decrease in receptor affinity (91, 92), and since the this cytokine pattern is observed in these patients (93), it is likely that cytokine-related mechanisms are involved in this process. It was subsequently reported that glucocorticoid resistant asthma was also associated with increased expression of the GRb isoform, suggesting that this splicing variant receptor might participate in the pathogenesis of glucocorticoid resistance in this condition (94). Since many phosphokinases and other molecules, which are important in cytokine and growth factor signaling, modulate GR activity (16, 18), it is possible that such molecules might also modulate receptor affinity in AIDS patients who also develop activation of the immune system and cytokine dysregulation.
An acquired form of lipodystrophy, which partially mimicked the clinical presentation of Cushing syndrome, was reported in AIDS patients (10, 55, 95-99). The patients had a characteristic redistribution of their adipose tissue, with an enlargement of their dorsocervical fat pad (“buffalo hump”), axial fat pads (bilateral symmetric lipomatosis), lipomastia, and expansion in their abdominal girth ("Crix-belly" or "protease paunch"), as well as thinning of the extremities and muscle wasting. Since all these manifestations are reminiscent of the typical phenotype of chronic glucocorticoid excess or Cushing syndrome, this condition was initially referred to as a pseudo-Cushing state, a term reserved for obese, depressive or alcoholic patients with biochemical hypercortisolism who are frequently hard to differentiate from true Cushing syndrome (100). In addition to its characteristic physical phenotype, the AIDS-related insulin resistance and lipodystrophy syndrome has a pathophysiologic and biochemical phenotype, which is reminiscent of that of Cushing syndrome, the metabolic syndrome, and some of the congenital lipodystrophy syndromes (101, 102). Thus, in this syndrome, we have an increased fat to lean body mass ratio, carbohydrate intolerance and/or diabetes mellitus, and dyslipidemia, all significant risk factors for cardiovascular disease.
What causes this recently recognized AIDS-associated syndrome? Possible mechanisms are those that cause or lead to insulin resistance and are listed in Table 1 and summarized in Figure 1. As several previous reports indicated, one of the earlier suggestions was that the syndrome was an adverse effect of protease inhibitors (103). These compounds efficiently inhibit the activity of the viral-encoded protease, which normally digests the Gag-Pol p160 kDa precursor protein, producing several polypeptide fragments with distinct functions. The catalytic site of the HIV-1 protease has about 60 % homology with the low-density lipoprotein receptor-like protein (LRP) and the C-terminal region of the cytoplasmic retinoic-acid binding protein type 1 (CRABP-1) (103). LRP plays an important role in triglyceride metabolism in adipocytes and hepatocytes by cooperating with lipoprotein lipase to cleave fatty acids from circulating triglycerides, thus allowing free fatty acids to enter into these tissues. CRABP-1 is a cytosolic retinoic acid-binding protein that controls the availability of this hormone inside the cells where retinoic acid plays an important role in growth, differentiation and apoptosis.
Table 1. Potential Contributing Factors in AIDS-related Insulin Resistance and Lipodystrophy before and after Treatment with Protease Inhibitors (Modified from references 97 and 98).
|
Before Rx |
After Rx |
|
|---|---|---|
|
+ = presence, - = absence, ? unknown, * During stress and starvation, both fat and lean body mass are lost. Post stress and starvation body weight gain is primarily due to fat accumulation. |
||
|
Nonspecific, disease-related |
||
|
Sickness-related starvation |
+ |
Refeeding |
|
Sickness-related change in body composition |
Lean body mass loss* |
Fat mass gain* |
|
Infection-induced hypercytokinemia |
+ |
- |
|
Cytokine-induced adipose tissue 11-hydroxysteroid dehydrogenase stimulation |
+ |
- |
|
Stress- and starvation-induced hypercortisolism |
+ |
- |
|
Specific, HIV-1-related |
||
|
Virally-induced muscle, liver, and fat glucocorticoid hypersensitivity |
+ |
? |
|
Virally-induced adipose tissue PPAR inhibition |
+ |
? |
|
Virally-induced adipose tissue 11-hydroxysteroid dehydrogenase stimulation |
+ |
? |
|
Protease-inhibitor-related |
||
|
Rx-induced-insulin resistance/dyslipidemia |
- |
+ |
|
Genetic/constitutional predisposition |
+ |
+ |
Figure 1. Major proposed mechanisms in the genesis of AIDS-related insulin resistance and lipodystrophy. Pre-therapy, insulin resistance may be induced by inflammatory cytokines, stress- and malnutrition- induced hypercortisolism, proteins of the HIV-1, such as Vpr, interfering with host cellular mechanisms, and loss of lean body mass. At the same time, glucocorticoid hypersensitivity and/or suppression of the PPRA activity may be induced by proteins of the HIV-1, such as Vpr and Tat, and by inflammatory cytokine-mediated stimulation of target tissue 11-hydroxysteroid dehydrogenase that converts inactive cortisone to active cortisol. Post therapy, insulin resistance may be induced by direct actions of protease inhibitors on insulin target organs, refeeding-related gain of mostly fat mass, and, possibly, proteins of the HIV-1, such as Vpr and Tat continuing to be expressed in insulin and glucocorticoid target organs (from references (101, 102, 133))
So far, no substantial binding of protease inhibitors to either of these molecules has been demonstrated, however, the hypothesis is attractive and testable. As a supporting evidence for this hypothesis, protease inhibitors were recently shown to modulate retinoic acid receptor activity, possibly through their effect on CRABP-1 (104). It was also reported that protease inhibitors suppressed the transport function of Glut4, the key trans-membrane glucose transporter that may lead to peripheral tissue resistance to insulin (105). In addition, protease inhibitors decreased hepatic lipase activity and modulated differentiation of pre-adipocytes, via unknown mechanisms that might also contribute to or exacerbate the clinical picture of AIDS-related insulin resistance and lipodystrophy (106-108). As a host genetic factor, an apo E polymorphism was associated with the dyslipidemia seen in AIDS patients treated with protease inhibitors (109).
Although the protease inhibitors are the most promising candidates for the causation of this syndrome, based on the evidence that the majority of AIDS patients develop this syndrome after taking such compounds, a small percentage of patients develop similar characteristic features prior to treatment or while on treatment with nucleoside-analogue reverse transcriptase inhibitors alone. This suggests that the HIV-1 infection itself could nonspecifically, -via inflammatory cytokine elevations and stress induced cortisol hypersecretion-, induce an insulin resistant phenotype vulnerable to protease inhibitors (110). Indeed, inflammatory cytokines, such as TNFa, IL-1 and IL-6, do cause resistance to insulin, both directly and via stimulation of the target tissue 11b-hydroxysteroid dehydrogenase of cortisone, which increases tissue exposure to cortisol. In this context, protease inhibitors might just exacerbate already present, smoldering insulin resistance and lipodystrophy, not expressed because of the known malnutrition of sick AIDS patients and the absence of sufficient calories to build visceral and other fat deposits (10, 101, 111).
As the manifestations of the sickness syndrome subside with treatment, the emaciated patient goes through refeeding with body weight gain of mostly fat, tilting the ratio of fat to lean body mass upward, further worsening insulin resistance.
HIV-1, in its 9.8 kb genomic information, encodes and produces 3 precursor proteins, the Gag, RNA polymerase and envelope polypeptides, whose processed products are reverse transcriptase, protease, integrase, matrix, and capsid, as well as 6 accessory proteins, Tat, Rev, Nef, Vif, Vpr and Vpu (112). Some of these polypeptides are virion-associated proteins incorporated into the viral particle and others are expressed in host cells where they direct viral replication and gene expression and several host cell functions. Since infection with HIV-1 has a dramatic impact on host target cells, it is quite possible that some of these viral proteins modulate host cell glucose and lipid metabolism, participating in the development of AIDS-related insulin resistance and lipodystrophy.
Since the clinical picture of this syndrome strongly overlaps with that of Cushing syndrome, stress-related hypercortisolism was originally hypothesized as a main potential causative factor in AIDS-related insulin resistance and lipodystrophy syndrome. We did examine the adrenal function of patients who developed this syndrome while treated with protease inhibitors, in parallel to that of patients with endogenous Cushing syndrome and revealed that AIDS-related insulin resistance and lipodystrophy had some distinct features from the glucocorticoid-induced condition (113). Thus, patients with this syndrome had normal plasma levels of basal and ovine CRH-stimulated ACTH and cortisol, and normal excretion of 24-hour urinary free cortisol. Moreover, their leukocyte GRs were in normal concentrations, while their affinity to dexamethasone was similar to that of controls (113). Their dyslipidemia was more severe than that of Cushing syndrome patients.
Therefore, biochemical hypercortisolism is not likely to be a major cause of AIDS-related insulin resistance and lipodystrophy in protease-treated AIDS patients. Rather, it is still possible that localized or tissue-specific hypersensitivity to glucocorticoids in key tissues like the adipose, skeletal muscle and liver may be involved. Indeed, there are several pieces of evidence indicating that AIDS patients have altered tissue sensitivity to glucocorticoids. First of all, they all develop reduction of innate and T helper 1-directed, cellular immunity. Levels of plasma IL-2, IL-12 and interferon-g, which direct cellular immunity, are suppressed in AIDS patients, while levels of IL-4 are increased (114, 115). All changes can be induced by exogenously introduced glucocorticoids and are seen in hypercortisolemic patients with classic Cushing syndrome (116). AIDS patients also frequently present with muscle wasting and myopathy, as well as dyslipidemia and visceral obesity-related insulin resistance (117-119). Therefore, it is possible that some unknown factor(s) might modulate tissue sensitivity to glucocorticoids in AIDS patients in a tissue-specific fashion, sparing their HPA axis, which is generally intact, suggesting preservation of normal negative feedback sensitivity to glucocorticoids.
In agreement with this finding, one of the HIV-1 proteins, Vpr, which is a 96-amino acid virion-associated accessory protein with multiple functions, including influencing transcriptional activity and having a cell cycle-arresting effect, increases the action of the GR by several fold, functioning as a nuclear receptor coactivator (120). The GR coactivator activity of Vpr is biologically evident in the suppression of IL-12 production from monocytes and the expression of activated NF-kB ligand (RANKL) in lymphocytes (121, 122). Vpr does this by cooperation with host cell coactivator integrator p300/CREB-binding protein (CBP) (121, 123, 124). These proteins regulate many signal transduction cascades mediated by transcription factors, including nuclear hormone receptors (NR), CRE-binding protein (CREB), activator protein-1 (AP-1), NF-kB and the signal transducers and activators of transcription (STATs) (125). p300/CBP have intrinsic histone acetyltransferase (HAT) activity and attract other histone acetyltransferase coactivators, such as the p300/CBP-interacting factor (p/CAF) and the p160-type nuclear receptor coactivator proteins to the transcription initiation site (125). Thus, p300/CBP act as regulatory “platforms” for the transcription of numerous host genes, regulating transcriptional activity of many trans-acting factors and NRs (Figure 2). Vpr contains a NR coactivator box similar to those present on p160 NR coactivators, binding to p300 and potentiating the effect of ligand-bound GR and p160. Vpr easily penetrates the cell membrane to exert its biologic effects even when added in the media surrounding the cells (126, 127), thus its effects may be extended to tissues not infected with HIV-1.
Figure 2. Linearized Vpr, Tat and p300 molecules and their mutual interaction domains. Vpr interacts with cellular molecules, such as NR, p300/CBP coactivators and 14-3-3, while Tat is physically associated with pTEFb elongation factor through its component Cyclin T1. Tat also binds p300/CBP and p160 type coactivators. Numerous transcription factors, transcriptional regulators and viral molecules bind the transcriptional coactivator p300. Binding sites of p160 nuclear receptor coactivators and Vpr overlap with each other and they both bind NRs and p300/CBP. Thus, Vpr mimics the host p160 nuclear receptor coactivators and enhances NR transcriptional activity. p300 facilitates attraction of transcription factors, cofactors and general transcription complexes by loosening the histone/DNA interaction through acetylation of histone tails by its histone acetyltransferase (HAT) domain. (modified from reference (101)). CREB: CRE-binding protein, HAT: histone acetyltransferase, NF-B: nuclear factor-B, NR: nuclear hormone receptor, p/CAF: p300/CBP-associating factor, pTEFb: positive-acting transcription elongation factor b, Rb: retinoblastoma protein, SF-1: steroidogenic factor-1, STAT2: signal transducer and activator of transcription 2, TFIIB: transcription factor IIB.
Another HIV-1 accessory protein, Tat, the most potent transactivator of the HIV-1-LTR, also moderately potentiates GR-induced transcriptional activity, possibly through accumulation of the positive-acting transcription elongation factor b (pTEFb) complex, that is comprised by the cyclin-dependent kinase 9 and its partner molecule cyclin T, on glucocorticoid responsive promoters (128). Because Tat, like Vpr, also circulates in blood and exerts its actions as an auto/paracrine or endocrine factor by penetrating the cell membrane (129), it is possible that it modulates tissue sensitivity to glucocorticoids irrespectively of a cell’s infection by HIV-1. Concomitantly with Vpr, Tat may induce tissue hypersensitivity to glucocorticoids that might contribute to viral proliferation indirectly, by suppressing local immune system activity and by altering the host’s metabolic balance, with both functions being governed by glucocorticoids (101, 102).
Vpr reduces tissue sensitivity to insulin not only through potentiating the actions of glucocorticoids, but also by modulating insulin’s transcriptional activity (101, 130, 131) (Figure 3). Insulin uses the forkhead transcription factors (FoxOs) to control gene induction; baseline unphosphorylated FoxOs are active, reside in the nucleus, and bind to their responsive sequences in the promoter region of insulin-responsive genes; in contrast, insulin activates Akt kinase, which phosphorylates specific serine and threonine residues of FoxOs rendering it inactive (132). Indeed, once FoxOs are phosphorylated at specific residues, they lose their transcriptional activity, by binding to and translocating into the cytoplasm with proteins of the 14-3-3 family (132). We found that Vpr moderately inhibited insulin-induced translocation of FoxO3a into the cytoplasm through inhibiting its association with 14-3-3 (131). Based on these in vitro findings, Vpr may be a key viral factor, which induces lipodystrophy, as well as insulin resistance and hyperlipidemia by interfering with and/or modulating cellular activities, such as transactivation of nuclear receptors or insulin (101, 133).
Figure 3. Vpr antagonizes the negative effect of insulin on FoxOs by retaining the latter in the nucleus via inhibition of its interaction with proteins 14-3-3. Insulin induces cytoplasmic translocation of FoxOs and inhibits their transcriptional activity by creating 14-3-3-binding sites through phosphorylation of these transcription factors. Vpr inhibits binding of 14-3-3 to phosphorylated FoxOs, thus it inhibits the negative effect of insulin on FoxOs, retaining them in the nucleus and constitutively stimulating the transcriptional activity of their responsive promoters. (from reference (101, 131))
We further found that this insulin resistance induction by Vpr might be compounded by the ability of the viral protein to interfere with the signal transduction of another nuclear receptor, peroxisome proliferation receptor (PPAR) g (126). Indeed, Vpr suppressed the c-Cbl associating protein (CAP) mRNA expression in pre-adipocyte cells and associated with the PPAR/RXR-binding site located in the promoter region of this gene. CAP is predominantly expressed in insulin-sensitive tissues and positively regulates insulin action, directly associating with both the insulin receptor and the c-Cbl proto-oncogene product (134). Vpr delivered either by exogenous expression or as protein added to media suppressed PPARg agonist-induced adipocyte differentiation, assessed as lipid accumulation and mRNA expression of the adipocyte differentiation marker aP2 in these cells (126). Thus, circulating Vpr, or alternatively Vpr produced as a consequence of direct infection of adipocytes, could suppress differentiation of preadipocytes by acting as a corepressor of PPARg-mediated gene transcription (101, 126). These results suggest that Vpr may alter sensitivity to insulin and thereby contribute to the development of lipodystrophy and insulin resistance observed in HIV-1-infected patients through suppression of PPARg activity, as well (101).
In summary, AIDS-related insulin resistance and lipodystrophy is most likely caused by multiple factors, including the infection itself, - via nonspecific inflammatory cytokine - and stress-induced hypercortisolism causing insulin resistance-, several HIV-1 products disturbing the cellular functions of the host, and viral protease inhibitors, all acting on a genetic and constitutional background of variable predisposition to the syndrome. Further studies are necessary to characterize this syndrome further, to better define the mechanisms involved in its development, and devise ways to prevent it from occurring or for reversing it.